Predictive maintenance made real with AI and cellular connectivity

Manufacturers need to reduce the time and effort required to have experts on-site for the maintenance of their equipment. This means finding the right balance for maintenance tasks, which frequently adds unnecessary costs, and increase the risk of failures. Introducing predictive maintenance, based on AI and cellular connectivity can address these issues. 

Predictive maintenance – the basics

Predictive maintenance is all about identifying problems before they occur, and the first step toward implementing it involves introducing sensors on equipment to measure anything measurable. Almost all new state-of-the-art industrial equipment already has embedded sensors. But existing machines can be retrofitted by simply attaching wireless sensors to them, and integrating the sensors with the factory IT/OT environment. All these sensors generate an immense amount of data- which can be analyzed by AI software, providing great insights that will help predict when maintenance is required to prevent potential faults.

Using data and AI to get the most out of factory equipment

The benefits of predictive maintenance are clear. It can reduce costs for bringing in expert technicians to maintain the equipment while preventing potential equipment failures. In a factory, when one machine is unexpectedly down, it can back up an entire assembly line, stopping productivity. A recent study from ABI Research found that one hour of downtime in an automotive factory can cost up to USD 30K per minute. Limiting the potential downtime on the shop floor can be directly translated onto the bottom line.

Making it real with cellular connectivity

For predictive maintenance to happen, highly reliable and secure wireless connectivity that supports a high density of devices is required. A 4G or 5G network infrastructure can fulfill these needs. For the enterprise environment, the network must be an easy-to-use private cellular network to deliver on the promise of security, reliability, and low latency.

It’s not just about the connectivity platform. The end-to-end solution requires devices, software, and connectivity to work together seamlessly. The ecosystem requires:

• Device and equipment vendors to provide products that connect with 4G and 5G technologies
• Software which can be easily integrated with solutions to collect, expose and manage the generated data
• System integrators with knowledge and capabilities to efficiently put everything together with a fast time to market.

We are actively working together with the broad ecosystem to bring all of this to life.

As an example, SAS Institute and Ericsson worked together to build a “predictive maintenance” proof of concept to perform real-time analytics based on measures taken from a running motor. These metrics, such as temperature and vibration, are registered from different wireless sensors that are connected to the cellular network and integrated into the factory IT environment. The results of the data analysis by the SAS software provide actionable feedback for a factory operator to keep the motor maintained, avoid potential issues, improve the downtime, and decrease costs.

Find more information about Connected Factory at SAS and about all of our partners in our Industry 4.0 site and our partner page.